TAKING A SPECTRUM PHOTOGRAPH.
A simple photograph may be obtained by the method illustrated in Fig. 762, re garding the real spectrum at s as the ob ject to be photographed ; a b are the parts of the spectrum under oi3servation, and s is the image of the spectrum. A quarter plate camera will be large enough for the first trial, and s may be focussed on the ground glass, either by means of the eye piece removing the camera lens, or rice wersti. In any case, see that the junc tion between the camera and the telescope tube is light-tight, and cover over the space between the lenses and with a thick black cloth. Focus first with a bright sodium light. Replace this by a limelight or an electric arc, and bring the part of the spectrum required on to the glass without altering the distance from L„. The spectrum may be an inch or more in length, from red to violet. The width is of no con sequence, and is generally kept small to prevent distortion of the lines. To get a satisfactory spectrum, however, the colours must be more dispersed. Addi
tional prisms, which must all be in mini mum deviation, are therefore introduced. Tn Fig. 762 two prisms P, are shown. Automatically acting tables are obtain able for keeping the prisms symmetrically apart, however the telescope may be moved. Without these, each prism has to be adjusted separately, beginning with the one nearest the collimator. As may be imagined, this multiplication of prisms means a large loss of light by reflection ; the source of light, therefore, must be good. The room should be darkened, the plate inserted in the camera in the usual way, and the exposure found by trial. For a small spectrum there will be an approximate focus all along the plate, but if only part of the spectrum be taken and the image consequently enlarged, the focussing screen should be stewed some what towards the violet, as the focal length of the lens for that colour will be shorter.